Closed-loop Deep Brain Stimulation, Synchrony breaking and Chimera State

Drexel University, Philadelphia PA

Investigators

Abstract

The primary goal of this project is to develop both computational and mathematical methods to break pathological synchrony and reduce bursting firing patterns presented in the basal ganglia of Parkinson' patients. A large-scale biologically faithful computational network is built to describe the parkinsonian state of the basal ganglia and thalamocortical circuit. In the computational method, the investigators conduct multi-site delayed feedback stimulation (MDFS) on the computational network. Various algorithms are developed to optimize the administration of MDFS. Based on the hypothesis that pathological outputs from the internal segment of the golobus pallidus (GPi) may induce parkinsonian signs by compromising thalamocortical (TC) relay, the objective function in the optimization is to diminish TC relay error in the parkinsonian state. In the mathematical method of synchrony breaking and burst reduction, a five dimensional system of nonlinear differential equations of GPi neuron is considered. A map reduction approach is applied to the GPi model neuron to obtain more tractable representation of the Poincare map that can capture the fundamental behavior of the GPi neuron. The reduced map is used to derive conditions that will settle a subgroup of uncoupled and fully synchronized GPi ensemble into different stable states of subthreshold oscillation, tonic fire and bursts with various spike numbers. The investigators expect to find the existence conditions of chimera states-the coexistence of synchrony and asynchrony-in this ensemble of GPi neurons that may lead to methods of mild stimulation of pathological synchrony breaking. The conventional deep brain stimulation delivers an ongoing stream of high frequency electrical pulses to a stimulation target in the brain. Even though the conventional deep brain stimulation has achieved remarkable successes in alleviating symptoms of Parkinson's patients, it has several significant drawbacks, such as high energy cost and laborious tuning in stimulation parameters. More importantly, such form of stimulation is considered "dumb" because the external stimulation force is not guided by the changes in the brain's electrical activity relevant to the disorder being treated. The computational frame work in this project will address the important challenges in advancing deep brain stimulation via less invasive and "smarter" closed loop protocols, multi-site delayed feedback stimulation, guided by the changes in the brain. The investigation of chimera states will link the phenomena of chimera states observed in dynamical systems to biophysically faithful models of neurological movement disorders. The successful finding of chimera states will assist in finding methods to break the synchrony of parkinsonian states with minimal amount of stimulation intervention.